AI in Healthcare
AI in Healthcare Research and Practice
Data and AI is one of the UK Government’s ‘Grand Challenges’ missions. It wants the UK to “Use data, Artificial Intelligence and innovation to transform the prevention, early diagnosis and treatment of chronic diseases by 2030”
Whilst the paradigm shift AI represents for healthcare research is not in doubt, the current consensus is that the best is still to come. The uses of Artificial Intelligence in a healthcare setting right now tend to be limited to “learning from patterns in the information the computer has used to learn, and therefore execute tasks”. In recent interviews carried out by LDA Research moderators, AI was not considered to be comparable, let alone a threat, to the accrued knowledge and experience of HCPs:
“I think AI, one of the things it could relieve me of are sort of tedious tasks that I have to do where I don’t add a whole lot of value, but I still need to do. It’s similar to lab tests that are commonly done, there’s not a laboratory physician sitting in the laboratory doing handheld blood counts, they’re done by an automated machine.” - US Radiologist
One interviewee felt that more AI technology may occasion the need for more HCPs to oversee it:
"One or two politicians seem to think that AI will solve radiology manpower shortages, and I think that’s an interesting point of view. I think it’s more likely that you will require more radiologists once you get more technology. You very often need more checks, more methods of interpretation, to make sure you get the right answer." - UK Radiologist
Current Uses of AI in Healthcare
Uses of AI technology in the UK are presently focused on automating time-consuming tasks in order to free up more time for HCPs to spend with their patients or improving their outcomes.
3 examples of AI in action demonstrate this:
- HeartFlow Cardiac Testing. The NHS is using this visualisation tech in place of angiograms. CT scans are used to create a 3D model of the heart with blood flowing around it. Doctors are able to see where blockages are disrupting the blood flow.
- DeepMind at Moorfields Eye Hospital. AI technology that identifies sight-threatening eye conditions at speed and is able to rank patients so that those in most urgent need are seen first.
- InnerEye Scan Processing. Used at Addenbrookes in Cambridge to speed up prostate cancer treatment. The AI scans images, outlines the prostate on the image, highlights tumours and presents a report.
Future Uses of AI in Healthcare Research
“The reality with AI is that once you’ve trained it, it should get better and better and better. That’s the way AI works. And so, how quickly do you adopt this?” UK Radiologist
The advantage AI has over the human brain is that it is able to observe and process vast amounts of data, whilst continuously improving the degree of accuracy with which it correlates learnt information. One area in which this is considered to be a game-changer is clinical research.
Recruiting for clinical trials is a costly activity, and AI solutions that minimise the cost are to be welcomed. Machine learning algorithms are being developed that can help researchers to recruit suitable candidates for trials by correlating and processing diverse data drawn from, for example, GP records, genetic information, and social media activity. This saves a huge amount of time and has the potential to deliver highly accurate results.
AI is also being developed that allows researchers to monitor participants more closely throughout trials. Real-time data access means that any adverse reactions, or biological changes can be picked up more quickly and dealt with swiftly.
The Challenges for AI in a Healthcare Setting
We are in the very early stages of AI adoption, and as the uses to which it can be put increase in their sophistication, so will the obstacles that need to be overcome:
- Transparency. This will need to be built into all AI applications, especially where medical procedures or products are being recommended. Doctors need to be able to see why a specific course of treatment is chosen.
- Privacy. There is huge public concern about private health data being accessed and used without permission. This makes accessing patient data difficult and time-consuming at present.
- Regulation. AI tech being developed for use in the European Economic Area (EEA) has to apply for a CE marking. It also has to meet the requirements of the EU Medical Device Regulation (MDR).
About LDA Research
LDA Research is a medical market research provider. Set up in 2011 by Lucy Doorbar, we specialise in providing global intelligence in the pharmaceutical industry and medical device sector. We have dedicated healthcare panels in the UK and US, as well as a network of international associates. The LDA team can be your eyes and ears wherever you need to be. From local culture to regulation and reimbursement, we can be relied upon to find the appropriate specialists for your requirements.